Two different questions
Sensitivity and Fundamentality concern the importance of variables and groups of variables, but they ask different questions. Ordinary sensitivity measures the association between variation in the sampled variables and variation in the response. Fundamentality asks how measured importance changes when a variable occupies a particular structural role in the sampled problem.
Matched samples
The current procedures distinguish a sample used for fundamentality from a separate matched sample used for ordinary sensitivity. Ordinary sensitivity should not be computed by simply reusing the fundamentality rows.
Absolute and relative Fundamentality
Individual absolute Fundamentality focuses on the variables designated as focal variables in the fundamentality rows; fixed or background coordinates are not treated as focal merely because they are present in the design description.
The individual absolute Fundamentality index and ordinary sensitivity use the same variance-ratio form but are evaluated on different matched samples. Relative Fundamentality is obtained by comparing the two; when ordinary sensitivity is zero or nonnumeric, the relative quantity is undefined.
Groups of variables
Group effects are evaluated through additive ANOVA comparisons between full and reduced models. The group sum of squares is the increase in residual error when the group is removed. The reported group index is normalized by total sum of squares; partial eta-squared and an F statistic are also available.
Outputs and modes
The procedures can produce separate CSV files for absolute Fundamentality, ordinary sensitivities, relative Fundamentality, summaries and group analyses. Modes allow complete runs as well as fundamentality-only, sensitivity-only and group-only calculations.
Interpretation
Fundamentality should be read as a measure of the structural dependence of measured importance within the specified sampling and problem representation, not as a universal intrinsic property of a variable.